Month: April 2018

I recently read comments on twitter about the story of the Speaker of the House firing the House’s chaplain. Apparently because the chaplain prayed against the recent tax bill. Or something.

It’s interesting to watch the perspective of the commenters. Assuming none of them were Russian trolls or George Soros paid actors, the perception of the appropriateness of the action clearly aligned with their political affiliation. This reminds me of people watching a professional sports game. Two spectators with roughly the same understanding of the game will interpret the actions of a referee/umpire/official very differently, based on which team he or she supports. An action against the opponent is reasonable and just, and an action against “their” team is unfair, or the result of malice or plain incompetence on the part of the official.

That’s really concerning to me. While the chaplain episode may not have much consequence in the larger scheme of things, applying this irrational reasoning to the political system will result in deeper divisions between parties and a willingness to “look the other way” by wrongdoing of the home team.

The political system appears to me to be built on the assumption that people are rational, but we are not.

The recent revelations about the Facebook / Cambridge Analytica debacle is alarming, but should not be surprising.  Companies, from Amazon to Netflix, continue to ratchet up the sophistication in how they analyze data to motivate potential customers.

It occurs to me, a political layperson, that the US electorate exists on a continuum: the far left and far right are highly aligned with their respective parties, often based on a few strong issues such as religion, abortion, taxes, gun rights, welfare, drug legalization, and so on.  But this population collectively likely only constitutes a minority of the electorate.  The majority reside in the middle.  The closer to the middle one gets, I hypothesize that there are an increasing number of issues that the one considers important in selecting a candidate to support.

Neither political party has to worry much about their “base”; it’s the group in the middle that they need to fight for.  From a lay person’s perspective, the approach candidates seem to generally take to address the “middle” is not the issues, but rather rote familiarity.  Basic marketing 101.  This is very likely why we see the costs of campaigns spiraling out of control: each candidate trying to get more “face time” in front of voters on TV, radio, and in front yards.  In a manner of speaking, there seemed to be tacit acceptance of this practice, because whomever could raise more money must, by definition, have more support as evidenced by their superior fund raising abilities.

I think the Facebook debacle (and likely the Trump presidency as a result) shows us that it’s possible to segment people in the “middle” and efficiently target them with specific messages that are intended to resonate with that particular group.

Given the plethora of data gathered on a continual basis about nearly all US citizens, it seems likely that elections going forward will become less about effectiveness of fund raising and more about the effectiveness of data analytics approaches and the ability to target and deliver specific messages to prospective voters that are intended to either want to vote for the desired candidate, or at least not vote for the opposing candidate.  All of this happens outside of public view and without any oversight.  And that should be troubling to everyone.

 

I’ve had a number of recent discussions regarding Georgia’s proposed new hacking law, as well as the proposed SESTA and FOSTA law.  Most of the people I’ve spoken to wonder why the laws would be proposed, since they will almost certainly cause the opposite effective of their intended purposes.  While this is certainly hard to objectively test, I strongly suspect a few things are happening here.

First, humans like certainty.  When we go to a doctor for some ailment, we want a doctor that can definitively tell us what the problem is.  If we go to a doctor with our ailment, and she tells us that she’s not sure what is wrong, we likely won’t go back.  Numerous studies have show, however, that doctors often misdiagnose illnesses, and that is one of the reasons machine learning and artificial intelligence holds great promise in medicine.  Regardless, we reward the doctor who is certain, and ignore that one who is uncertain.

I believe the same is true of politicians.  Like doctors, we, at least in the USA, have little patience for politicians who do not have robust positions on various topics.  Further, and unlike doctors, politicians are almost never allowed to change their minds on an issue, based on new input.  Mind-changing by a politician is almost always seized on by a political adversary as a sign of weakness.  Recall John Kerry’s debacle of voting for something before he voted against it.

Next, as Nassim Taleb points out, people are susceptible to what he calls the “narrative fallacy”.  People can be given a tangible narrative that makes intuitive sense, and in their minds, that narrative effectively becomes fact.  This is troublesome in the hands of a politician who is unwilling to change her mind based on new evidence.

Finally, and related to the narrative fallacy, politicians seem to have much greater success in selling policy platforms to the masses that are based on narratives, particularly narratives that are based on “fairness”, than it is to sell a position that may be counter-intuitive, but based on data.  Add in lobbying from firms that benefit from the incorrect narrative, and I suspect we end up with laws intended to do one thing, but end up causing the opposite.

In the case of SESTA and FOSTA, intended to drive down instances of sex slavery, the law may end up hurting more people than it can help.  There is a clear and straight-forward narrative for banning sex trafficking on the Internet, and holding those who provide aid, such as web site operators, accountable for their role in the crime.  Opponents point to data that indicate the opposite will happen, because the crimes will move “underground”, since demand has not changed, and those that were making a living off of sex work will have to resort to potentially much more dangerous avenues for earning money, such as by working through often violent pimps.

To counteract this phenomenon, we need to change our values around.  We should value those who do update their views based on new data.  We should understand that some issues are complex and expressing uncertainty is not weakness.  We should value people that want to gather additional information before making significant decisions, such as passing some new law.  We should strongly consider putting our senses of “fairness” in check when supporting or opposing positions that, on the surface may seem more fair, but in the end, cost everyone considerably more.

All of that is difficult to do because we are emotional beings, reacting to things through the lenses of the many different heuristics we’ve developed over thousands of millennia.  What we think is a rational position may not be rational, and we should not only be willing to accept that we’re wrong, but seek evidence that we are wrong and make changes accordingly.

In this blog, I intend to write about non-security related topics that interest me.  A bit about me:

• I work as a cyber security executive at a large company.
• I founded and co-host the Defensive Security Podcast
• I manage the infosec.exchange Mastodon instance
• I write about cyber security on my blog at infosec.engineering.
• I am fascinated by human cognition and the impacts that various heuristics and biases have on our daily lives.  I am particularly interested in how humans (mis)perceive risk.
• I am an avid audio book listener.
• I’m @maliciouslink on Twitter where an scary number of people follow me, mostly because I post pictures of my dog.

Welcome again, and thanks for visiting.